Apoptosis, or programmed cell death, plays a critical role in the development and maintenance of homeostasis in all multicellular organisms. Susceptibility to apoptosis varies markedly among cells and is influenced by both external and internal cellular events. Positive and negative regulator proteins that mediate cell fate have been defined, and dysregulation of these protein signaling networks has been documented in the pathogenesis of a wide spectrum of human diseases, including a variety of cancers. BCL-2 is the founding member of this family of apoptotic proteins and was first identified at the chromosomal breakpoint of t(14; 18)(q32; q21) lymphomas (Bakhashi et al. 1985 Cell 41:899; Cleary et al. 1985 Proc. Nat'l. Acad. Sci. USA 82:7439).
Gene rearrangement places BCL-2 under the transcriptional control of the immunoglobulin heavy chain locus, generating inappropriately high levels of BCL-2 and resultant pathologic cell survival. Such aberrations in apoptosis have been identified in lymphocytic and myelogenous leukemias and a host of other malignancies, and have been linked to tumor progression and acquired resistance to chemotherapy-induced apoptosis. The BCL-2 family of proteins has expanded significantly and includes both pro- and anti-apoptotic molecules that provide the checks and balances that govern susceptibility to cell death (FIG. 1). Not surprisingly, apoptotic proteins have become key targets for the development of therapeutics to both prevent precipitous cell death in diseases of cell loss and activate cell death pathways in malignancy.
The BCL-2 family is defined by the presence of up to four conserved “BCL-2 homology” (BH) domains designated BH1, BH2, BH3, and BH4, all of which include α-helical segments (Chittenden et al. 1995 EMBO 14:5589; Wang et al. 1996 Genes Dev. 10:2859). Anti-apoptotic proteins, such as BCL-2 and BCL-XL, display sequence conservation in all BH domains. Pro-apoptotic proteins are divided into “multidomain” members (e.g. BAK, BAX), which possess homology in the BH1, BH2, and BH3 domains, and the “BH3-domain only” members (e.g. BID, BAD, BIM, BIK, NOXA, PUMA), that contain sequence homology exclusively in the BH3 amphipathic α-helical segment. BCL-2 family members have the capacity to form homo- and heterodimers, suggesting that competitive binding and the ratio between pro- and anti-apoptotic protein levels dictates susceptibility to death stimuli. Anti-apoptotic proteins function to protect cells from pro-apoptotic excess, i.e., excessive programmed cell death. Additional “security” measures include regulating transcription of pro-apoptotic proteins and maintaining them as inactive conformers, requiring either proteolytic activation, dephosphorylation, or ligand-induced conformational change to activate pro-death functions. In certain cell types, death signals received at the plasma membrane trigger apoptosis via a mitochondrial pathway (FIG. 2). The mitochondria can serve as a gatekeeper of cell death by sequestering cytochrome c, a critical component of a cytosolic complex which activates caspase 9, leading to fatal downstream proteolytic events. Multidomain proteins such as BCL-2/BCL-XL and BAK/BAX play dueling roles of guardian and executioner at the mitochondrial membrane, with their activities further regulated by upstream BH3-only members of the BCL-2 family. For example, BID is a member of the “BH3-domain only” subset of pro-apoptotic proteins, and transmits death signals received at the plasma membrane to effector pro-apoptotic proteins at the mitochondrial membrane. BID has the unique capability of interacting with both pro- and anti-apoptotic proteins, and upon activation by caspase 8, triggers cytochrome c release and mitochondrial apoptosis. Deletion and mutagenesis studies determined that the amphipathic α-helical BH3 segment of pro-apoptotic family members functions as a death domain and thus represents a critical structural motif for interacting with multidomain apoptotic proteins. Structural studies have demonstrated that the BH3 helix interacts with anti-apoptotic proteins by inserting into a hydrophobic groove formed by the interface of BH1, 2 and 3 domains. Activated BID can be bound and sequested by anti-apoptotic proteins (e.g., BCL-2 and BCL-XL) and can trigger activation of the pro-apoptotic proteins BAX and BAK, leading to cytochrome c release and a mitochondrial apoptosis program.
BAD is also a “BH3-domain only” pro-apoptotic family member whose expression likewise triggers the activation of BAX/BAK. In contrast to BID, however, BAD displays preferential binding to anti-apoptotic members, BCL-2 and BCL-XL. Whereas the BAD BH3 domain exhibits high affinity binding to BCL-2, BAD BH3 peptide is unable to activate cytochrome c release from mitochondria in vitro, suggesting that BAD is not a direct activator of BAX/BAK. Mitochondria that overexpress BCL-2 are resistant to BID-induced cytochrome c release, but co-treatment with BAD can restore BID sensitivity. Induction of mitochondrial apoptosis by BAD appears to result from either: (1) displacement of BAX/BAK activators, such as BID and BID-like proteins, from the BCL-2/BCL-XL binding pocket, or (2) selective occupation of the BCL-2/BCL-XL binding pocket by BAD to prevent sequestration of BID-like proteins by anti-apoptotic proteins. Thus, two classes of “BH3-domain only” proteins have emerged, BID-like proteins that directly activate mitochondrial apoptosis, and BAD-like proteins, that have the capacity to sensitize mitochondria to BID-like pro-apoptotics by occupying the binding pockets of multidomain anti-apoptotic proteins.
The objective of identifying or generating small molecules to probe apoptotic protein functions in vitro and specifically manipulate apoptotic pathways in vivo has been challenging. High throughput screening has identified several molecules that inhibit the interaction of the BAK BH3 domain with BCL-XL at micromolar affinities. In addition to the potential drawback of identifying low affinity compounds, the technique is limited in its ability to generate panels of compounds tailored to the subtle binding specificities of individual members of protein families. Alternate approaches to manipulating apoptosis pathways have derived from peptide engineering, a technique that uses non-specific peptide sequence to generate compounds with desired three-dimensional structures. One application of this technique involved the generation of “pro-apoptotic” α-helices comprised of nonspecific peptide sequence used to induce cell death by disrupting mitochondrial membranes.
The alpha-helix is one of the major structural components of proteins and is often found at the interface of protein contacts, participating in a wide variety of intermolecular biological recognition events. Theoretically, helical peptides, such as the BH3 helix, could be used to selectively interfere with or stabilize protein-protein interactions, and thereby manipulate physiologic processes. However, biologically active helical motifs within proteins typically have little structure when taken out of the context of the full-length protein and placed in solution. Thus, the efficacy of peptide fragments of proteins as in vivo reagents has been compromised by loss of helical secondary structure, susceptibility to proteolytic degradation, and inability to penetrate intact cells. Whereas several approaches to covalent helix stabilization have been reported, most methodologies involve polar and/or labile crosslinks (Phelan et al. 1997 J. Am. Chem. Soc. 119:455; Leuc et al. 2003 Proc. Nat'l. Acad. Sci. USA 100:11273; Bracken et al., 1994 J. Am. Chem. Soc. 116:6432; Yan et al. 2004 Bioorg. Med. Chem. 14:1403). Subsequently, Verdine and colleagues developed an alternate metathesis-based approach, which employed α,α-disubstituted non-natural amino acids containing alkyl tethers (Schafmeister et al., 2000 J. Am. Chem. Soc. 122:5891; Blackwell et al. 1994 Angew Chem. Int. Ed. 37:3281).